In the early years when the cost of feeding a child is small, growth must depend largely on the patience and conscientiousness of the mother and on the adequacy of her knowledge. And we have seen that the standard of maternal care declines with increasing family size, in the poorer groups especially. At later ages the cost of the food itself becomes important, and the poor growth of the later-born child will be due either to an inadequate family income or unwise spending. The marked relation between the standard of maternal care and growth in those families whose income, though not large, should be sufficient-suggests that spending habits are important. But the lack of such a relationship in the large families of skilled manual workers, and in all families in the poorest groups, suggests that below a certain level of income even the most careful spending will not provide a diet fully adequate for growth. (33)
It seems fair to say that in their families, as in those in Candelaria, family size could be termed an aggravating factor.
By this rather cumbersome term is meant the effects associated with family size which remain after other factors considered causal are controlled for. The data need not be reviewed in detail here, but several of the studies described earlier presented relevant evidence and three are worth mentioning specifically.
Morris and his associates found that when they controlled for social class and maternal age, infant mortality increased with parity and the increase was relatively greatest in the younger mothers in the highest social classes. (See Figure 3.) Whatever benefits one might expect from improved environmental circumstances associated with higher social classes are offset by family size. This is not to say, however, that family size is the only residual variable remaining to account for variations in mortality. Maternal age is also involved, and this prompts a consideration of maternal competence, since it is the younger mothers who have the problem. These two factors, and probably others, are almost certainly interacting.
The data from Douglas and his colleagues also showed clear-cut effects on growth and intelligence associated with family size when social class was held constant. What is clearly evident from their data (see Table 5) is that children from small families in the lowest social classes grow as well as their age peers in the higher social classes. In other words, lower-class mothers can and do overcome, or compensate for, whatever factors are operating to interfere with growth and development, providing the pressure of numbers is not too great. On the other hand, family size has no effect in the higher social classes.
This is not the case, however, with regard to intelligence. Table 9 shows that test scores are higher in the smallest families, but social class differences